A STUDY DESIGNED TO IDENTIFY SOME OF THE DIFFERENCES BETWEEN THE RESPONSES ON THE GATES ADVANCED PRIMARY READING TEST AND THE KINDS OF RESPONSES OBTAINED FROM AN INFORMAL READING INVENTORY (IRI) IS REPORTED. SUBJECTS WERE 65 THIRD-GRADE PUPILS IN WEST BABYLON, NEW YORK. PUPILS AT THE SAME INSTRUCTIONAL LEVEL SCORED HIGHER IN THE RECOGNITION TEST (GATES) THAN IN THE RECALL TEST (IRI). THE RESPONSES OF THE PUPILS ILLUSTRATED THAT THE RESULTS OF A TEST SUCH AS THE GATES CANNOT BE ANALYZED FULLY UNLESS THE MANNER IN WHICH THE RESPONSES WERE DERIVED CAN BE DETERMINED. ON THE IRI, WHERE THE MAJORITY OF TESTING IS ORAL, GUESSING IS HELD TO A MINIMUM. AN ANALYSIS OF PARTIAL READING INDICATED THAT PUPILS CAN READ INACCURATELY AND STILL OBTAIN THE CORRECT ANSWER TO ITEMS ON A STANDARDIZED TEST SUCH AS THE GATES. THEREFORE, INACCURATE READING ALONE DOES NOT ACCOUNT FOR CORRECT OR INCORRECT ANSWERS ON THAT TEST. SIGNIFICANT DIFFERENCES BETWEEN THE TESTS INDICATE THAT THE GATES GRADE-PLACEMENT SCORE REFLECTS A MORE GLOBAL MEASURE OF READING PERFORMANCE THAN DOES THE INSTRUCTIONAL LEVEL OF THE IRI AND THAT THE GATES SHOULD NOT BE EXPECTED TO GIVE A SCORE EQUIVALENT TO THE INSTRUCTIONAL READING LEVEL OF THE INFORMAL TEST. TABLES AND REFERENCES ARE GIVEN. THIS PAPER WAS PRESENTED AT THE INTERNATIONAL READING ASSOCIATION CONFERENCE (SEATTLE, MAY 4-6, 1967). (RH)